Abstract Measurement and mechanical aspects of the microbond pull-out technique for measuring fiber/resin interfacial shear strength have been examined to determine the origins of the consistent variability in bond strength results. Improved instrumentation for measuring fiber diameters and debonding loads make it possible to rule out these measurements as a significant cause of bond strength variability. Bond failure under the microbond shearing conditions is catastrophic. Variations in droplet contour near its intersection with the fiber are a consequence of the variations of contact angle between the two surfaces. Since the contact angle is less than 90°, a compressive force which acts against the shearing process must be generated, but the variability of bond strength values cannot be accounted for by this contact angle-dependent compressive loading. Shear stress distributions along the fiber/droplet interface and contact angle variations make it vital that the shearing plates be as close as possible to the intersection of the fiber and resin surfaces. The results presented in this paper reinforce the idea that the principal cause of bond strength measurement variability lies in the inherent nonuniformity of fiber surfaces.